Number of sig figs in traits and other tables

[serbinsh]

I noticed today that some trait data has unreasonable sig figs, e.g.

Does this matter or would it be better to limit this in the data tables? I assume that this could help reduce storage requirements, though they arent really that high yet. Perhaps not an important issue right now?

[mdietze]

Storage costs are determined by the schema, not the number (e.g. if it's a double, then 0 takes up just as much storage as any other number). The origin of such numbers is another issue, as is how we display them in the web interface

[gsrohde]

My guess is that whoever entered these numbers was given numbers with a only a few s.d., converted to the units used in BETYdb, and just kept all of the digits displayed in their calculator.